Skip to main content

Table 4 Power calculation for changes in practice for various assumptions of % with practice pre-intervention – p1(20–60%), p2 (30–90%), assuming 80% power at 5% level of statistical significance (2-sided test), for sizes of effect ≤ 2 and for cluster size six (numbers of annual live births per hospital)

From: The BLISS cluster randomised controlled trial of the effect of 'active dissemination of information' on standards of care for premature babies in England (BEADI) study protocol [ISRCTN89683698]

ICC assumption required for 780 babies
p 1 P 2
  30 35 40 45 50 55 60 65 70 75 80 85 90  
60          0.02 0.31     
55          0.28      
50         0.26       
45        0.25 0.6       
40       0.25 0.6        
35     0.01 0.26 0.6         
30    0.02 0.28           
25   0.04 0.31            
20 0.06 0.36